Brown, S.C. , Mason, C.A. , Lombard, J.L. , Martinez, F., Plater-Zyberk, E., Spokane, A.R. , Newman, F.L. , Pantin, H., & Szapocznik, J. (2009). The relationship of built environment to perceived social support and psychological distress in Hispanic elders: The role of “ Eyes on the Street. ” Journal of Gerontology: Social Sciences, 64B(2), 234–246, doi:10.1093/geronb/gbn011, Advance Access publication on February 4, 2009

© The Author 2009. Published by Oxford University Press on behalf of The Gerontological Society of America.All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

234

STUDIES of the relationship between the physical char-acteristics of neighborhoods and residents ’ health and

behavior (e.g., cardiovascular disease, depression, physical activity) have increased appreciably in recent years (for re-views, see Diez Roux, 2003 ; Evans, 2003 ; Frumkin, Frank, & Jackson, 2004 ; Institute of Medicine and Transportation Research Board, 2005 ; Northridge, Sclar, & Biswas, 2003 ). Concepts such as mixed land use and walkability at the neighborhood level have been shown to be related to residents ’ health and social well-being ( Clarke & George, 2005 ; Frank, Andresen, & Schmid, 2004 ; Kim & Kaplan, 2004 ; Leyden, 2003 ; Rohrer, Pierce, & Denison, 2004 ; Szapocznik et al., 2006 ). In contrast to the growing body of research relating built environment to health/wellness, few studies have examined the relationship of the neighborhood built environment to elders ’ mental health (e.g., Berke, Gottlieb, Vernez Moudon, & Larson, 2007 ).

The present study investigates a further construct of “ eyes on the street ” ( Jacobs, 1992 ), which involves the use of ar-chitectural and neighborhood design features that promote direct observation and interaction among individuals in a building and individuals walking in the street ( Jacobs, 1992 ; Katz, 1994 ; Leccese & McCormick, 2000 ). Architectural features such as porches, stoops, and windows are hypoth-esized to facilitate social interactions and monitoring of behavior among residents ( Evans, 2003 ; Leccese & McCormick, 2000 ; Leyden, 2003 ), which in turn promote social support ( Jacobs, 1992 ; Kawachi, 1999 ; Leccese & McCormick, 2000 ). Such outcomes are highly desirable, given the strong evidence that social support is associated with a variety of positive health and behavioral health out-comes ( Adler & Matthews, 1994 ; Cohen, 2004 ; House, Landis, & Umberson, 1988 ; Newsom & Schulz, 1996 ). The present study examined the built environment’s impacts on

The Relationship of Built Environment to Perceived Social Support and Psychological Distress in Hispanic

Elders: The Role of “ Eyes on the Street ”

Scott C. Brown , 1 Craig A. Mason , 2 Joanna L. Lombard , 1 , 3 Frank Martinez , 1 , 3 Elizabeth Plater-Zyberk , 3 Arnold R. Spokane , 4 Frederick L. Newman , 5 Hilda Pantin , 1 and José Szapocznik 1 , 3

1 Department of Epidemiology and Public Health, Center for Family Studies, University of Miami, Miller School of Medicine, Florida.

2 University of Maine, College of Education and Human Development/Center for Community Inclusion and Disability Studies, Orono.

3 University of Miami School of Architecture, Coral Gables, Florida. 4 Department of Education and Human Services, Lehigh University College of Education, Bethlehem, Pennsylvania.

5 Robert Stempel School of Public Health, Department of Health Policy and Management, Florida International University, Miami.

Background. Research on contextual and neighborhood effects increasingly includes the built (physical) environ-ment’s infl uences on health and social well-being. A population-based study examined whether architectural features of the built environment theorized to promote observations and social interactions (e.g., porches, windows) predict Hispanic elders ’ psychological distress.

Methods. Coding of built environment features of all 3,857 lots across 403 blocks in East Little Havana, Florida, and enumeration of elders in 16,000 households was followed by assessments of perceived social support and psychological distress in a representative sample of 273 low socioeconomic status (SES) Hispanic elders. Structural-equation modeling was used to assess relationships between block-level built environment features, elders ’ perceived social support, and psychological distress.

Results. Architectural features of the front entrance such as porches that promote visibility from a building’s exterior were positively associated with perceived social support. In contrast, architectural features such as window areas that promote visibility from a building’s interior were negatively associated with perceived social support. Perceived social support in turn was associated with reduced psychological distress after controlling for demographics. Additionally, perceived social support mediated the relationship of built environment variables to psychological distress.

Conclusions. Architectural features that facilitate direct, in-person interactions may be benefi cial for Hispanic elders ’ mental health.

Key Words: Aging — Built environment — Hispanics/Latinos — Psychological distress — Social support .

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 235

community-dwelling urban elders ’ social support, on which elders may be especially dependent ( Cantor, 1979 ; Shaw, 2005 ; Thompson & Krause, 1998 ; Wethington & Kavey, 2000 ), which may in turn impact their psychological dis-tress ( Cornman, Goldman, Glei, Weinstein, & Chang, 2003 ; Krause, Liang, & Yatomi, 1989 ; Newsom & Schulz, 1996 ).

Built Environment and Social Support A limited but growing body of research has supported a

possible connection between features of the built environ-ment and social behavior. Aspects of the built environment, such as semiprivate space, seating, and open space, have been found to be related to increased supportive acts by neighbors and a higher sense of community attachment ( Skjaeveland & Garling, 1997 ). Similarly, housing patterns that contain clustered homes with a shared central space have been found to increase spontaneous social interactions and connectedness with neighbors ( Abu-Ghazzeh, 1999 ). Other work suggests that characteristics of outdoor common spaces may play a role in the formation and maintenance of social ties in inner-city elders ( Kweon, Sullivan, & Wiley, 1998 ). Such observations are signifi cant because although research suggests that social support can be obtained with-out direct contact, face-to-face interactions are associated with greater perceived availability and adequacy of support ( Seeman & Berkman, 1988 ; Wellman et al., 1973 ). Conse-quently, if physical characteristics of neighborhoods are not conducive to establishing contact, then obtaining support may be more challenging, particularly in older residents ( Thompson & Krause, 1998 ).

Relevant to such research, urbanist theories such as those proposed by Jacobs (1992) and New Urbanist theory ( Leccese & McCormick, 2000 ) hold that positioning buildings with porches, stoops, and windows close to the street or other public space promotes surveillance of the surrounding neighborhood and also allows neighbors to look out for one another ( Jacobs ; Leccese & McCormick, 2000 ). Combined with functional opportunities for interaction, such as resi-dential entrances near pedestrian pathways, these features may promote increased social interaction and heightened social support ( Evans, 2003 ; Fleming, Baum, & Singer, 1985 ; McCoy, 2002 ; Moos, 1976 ).

Social Support and Psychological Distress The relationship between social support and both health

and behavioral health (e.g., immune function, depression) is well established ( Adler & Matthews, 1994 ; Cohen, 2004 ; House et al., 1988 ; Newsom & Schulz, 1996 ), with evidence that the benefi ts associated with social support and social interactions are particularly important for the elderly ( Cornman et al., 2003 ; Krause et al., 1989 ; Liang, Krause, & Bennett, 2001 ; Lubben, 1988 ; Newsom & Schulz, 1996 ; Shaw, 2005 ). There is a particularly strong link between social

support and depressive symptoms ( Chou, 1999 ; Cohen & Wills, 1985 ; Newsom & Schulz, 1996 ). Moreover, per-ceived social support in particular may reduce older adults ’ depression or anxiety over time because believing that emo-tional or tangible support is available if needed may cause people to spend less time worrying about life’s problems and daily hassles ( Peirce, Frone, Russell, Cooper, & Mudar, 2000 ). In fact, prior research suggests that subjective evalu-ations of supportive encounters may be more strongly re-lated to mental health than are objective markers of social support (e.g., frequency of contact with others; Jang, Haley, Small, & Mortimer, 2002 ; Krause, 1995 ). In addition, per-ceived support may enhance emotional functioning by pro-moting less threatening interpretations of adverse events and more effective coping strategies ( Cohen & Wills, 1985 ). Further, satisfaction with one’s social support has been shown to be inversely related to future depression in older adults ( Krause et al., 1989 ).

Built Environment, Social Support, and Psychological Distress

In summary, the existing literature suggests a model in which social support mediates the impacts of the built envi-ronment on elderly residents ’ health and mental health out-comes. For example, the fi nding that older adults who reside in physically deteriorated neighborhoods perceive that so-cial support is less available to them ( Thompson & Krause, 1998 ) and have worse health ( Krause, 1996 , 1998 ) than do elders who reside in better-maintained neighborhoods sug-gests that low social support may partly account for the re-duced health ( Krause, 1996 ; Thompson & Krause, 1998 ) and possibly mental health ( Krause, 1993 ) of elders resid-ing in physically dilapidated neighborhoods. Similarly, fea-tures of the neighborhood built environment (e.g., physical dilapidation, neighborhood walkability) have been shown to predict individuals ’ levels of psychological distress even after controlling for individual-level variables such as age, gender, and fi nancial strain ( Berke et al., 2007 ; Krause, 1996 ; Kruger, Reischl, & Gee, 2007 ; Weich et al., 2002 ). Furthermore, the ability of neighborhood features to pro-mote social ties and social support has been hypothesized as one mechanism through which the built environment may operate on psychological distress ( Berke et al., 2007 ; Evans, 2003 ; Kruger et al., 2007 ; Weich et al., 2002 ). We further hypothesize that the built environment’s effects on elderly residents ’ psychological distress will be mediated by elders ’ perceptions of the quantity and quality of the social support that they receive. However, to our knowledge few studies have examined these relationships simultaneously in any population.

Proposed Model Based on the literature, the current study tested a two-step

sociocontextual model examining (a) whether built

BROWN ET AL. 236

environment ‘ eyes on the street ’ features theorized to pro-mote observation and interaction are associated with elders ’ perceived social support and (b) whether perceived social support in turn is associated with reductions in psychological distress. We further examined whether perceived social sup-port mediated the relationship between the built environment and elders ’ psychological distress. To our knowledge, this is the fi rst study that evaluates prospectively the impact of the built environment on these social environment and psy-chological variables in a single model. Moreover, we ex-amined these relationships in a population-based sample of urban Hispanic elders of low socioeconomic status (SES), in whom we previously found a high rate (35%) of clinically relevant depressive symptoms ( Perrino, Brown, Mason, & Szapocznik, in press ): This population may be at particular risk for psychological distress due to a variety of factors, includ- ing limited fi nancial circumstances, social isolation, and resi-dence in inner-city neighborhoods ( Falcón & Tucker, 2000 ; Perrino et al., in press ; Stewart & Nápoles-Springer, 2000 ). We therefore examined the possible relationship of contex-tual variables (i.e., neighborhood built environment and el-ders ’ perceived social support) to psychological distress in this study population, after controlling for demographics.

Methods

Setting The study was conducted in East Little Havana, Florida,

a well-defi ned urban community within the City of Miami, which in 2002 was the poorest large city in the United States ( U.S. Census Bureau, 2002 ). East Little Havana is 93% His-panic, with 19% of residents 65 years of age or older and 35% of residents living below the poverty level ( U.S. Census Bureau, 2000 ). This community includes 3,857 lots in 403 blocks. East Little Havana is relatively densely built, with 40,865 residents living in an area of less than 3 square miles ( U.S. Census Bureau, 2000 ). East Little Havana was selected for examination because it is a relatively homogeneous neighborhood with respect to sociodemographic character-istics (i.e., mostly Hispanic, with the highest rate of poverty in the county; U.S. Census Bureau, 2000 ) and yet demon-strates considerable variability in the built environment. For instance, built environment features, such as porches or win-dows, which were theorized to facilitate observation and so-cial interaction ( Jacobs, 1992 ; Leccese & McCormick, 2000 ), were differentially present across the blocks of this neighborhood, ranging from blocks without any porches or windows facing the street to blocks with many porches and windows facing the street ( Spokane et al., 2007 ).

Research Design This study is part of a larger, population-based, prospec-

tive cohort study entitled “ The Hispanic Elders ’ Behavioral Health Study, ” which examines the relationship of the

neighborhood built and social environment to Hispanic el-ders ’ mental health outcomes. This study was approved by the University of Miami’s Institutional Review Board. To obtain data on the built environment, the 3,857 lots in the 403 blocks in East Little Havana were each assessed for their built environment features between 2000 and 2002. To obtain a population-based sample of Hispanic elders, a door-to-door survey enumerated all 16,000 households in the community, identifying 3,322 community-dwelling His-panic elders 70 years of age or older living on 302 blocks. Eligible elders were located by trained Spanish-speaking enumerators who made up to 11 home visits to each address at different times of day and on multiple days of the week to identify the eligible elders in the neighborhood.

A sample of elders ages 70 and above was selected to maximize the likelihood of identifying an effect of the built environment on social and behavioral health outcomes. Horgas, Wilms, and Baltes (1998) found that elders ages 70 and above spend approximately 80% of their time at home in their neighborhoods. One Hispanic elder was ran-domly selected from each of the 302 blocks on which el-ders lived. If an elder refused to participate or did not meet inclusion criteria, a second randomly selected elder was approached, and so on, until one elder in each of the blocks with elders had been consented. Through this process, 521 elders were ultimately approached for possible participa-tion, but of this total 248 were lost for the following rea-sons: 30 had died since enumeration, 95 refused, 87 had moved or could not be located/contacted after 11 home vis-its, 10 had incorrect home addresses, 24 did not meet other eligibility criteria, and 2 moved to a block from which an elder had already been sampled. The fi nal sample at base-line consisted of 273 eligible elder participants living one each in 273 blocks (i.e., we were unable to sample elders from 29 of the 302 blocks on which elders resided, despite random resamplings from the same block, which occurred primarily because most of these 29 blocks had only one or two potential elders residing on them). As part of the larger study, participants completed four annual assessments of social support and behavioral health (i.e., psychological distress and cognition). The present analyses utilize base-line assessment data, with most baseline assessments com-pleted in 2002 – 2003.

Participants Inclusion/exclusion criteria were (a) 70 years of age or

older, (b) born in a Spanish-speaking country, (c) resident of East Little Havana, (d) lives in housing in which he/she can walk outside (this excludes nursing homes and specialized locked housing units), (e) of suffi cient physical health to go outside, and (f) obtained a score of 17 or above on the Mini-Mental State Examination (MMSE), a global measure of cognitive functioning ( Folstein, Folstein, & McHugh, 1975 ). The fi nal sample was 59% women and 87% Cuban born,

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 237

with a mean age of 78.5 years ( SD = 6.3), an average of 7.3 years of education ( SD = 4.3), and a mean annual household income of $9,300 ( SD = $4,550). Participants ’ average score on the MMSE was 24.8 ( SD = 3.2). Thirty-four percent of participants were married and 7% were employed at base-line. Participants reported living in their home for an aver-age of 13.7 years ( SD = 11.3) and with an average household size of 1.8 residents ( SD = 1.1). Participants were paid $25 for the baseline interview.

Measures Control variables: Control variables included household

income, gender (0 = men, 1 = women), age, and education in years. In addition, we also controlled for global cognitive functioning, as measured by the MMSE ( Folstein et al., 1975 ), given prior fi ndings that cognitive functioning is as-sociated with social functioning and social support in older adults (e.g., Plehn, Marcopulos, & McLain, 2004 ; Zimmer, Ofstedal, & Chang, 2001 ).

Built environment: The built environment of the study area was measured using the University of Miami Built Environ-ment Coding System (UMBECS; Spokane et al., 2007 ), a comprehensive coding system developed to assess built envi-ronment characteristics ( Duany, 2000 ; Leccese & McCormick, 2000 ; Lombard et al., 2000 ; Spokane et al., 2007 ). Architec-ture students, trained using the UMBECS, coded each of the 3,857 lots in East Little Havana. Interrater reliabilities of .80 were required for a student to serve as a rater and for her/him to continue to conduct ratings throughout the study. As part of a larger assessment, each lot was coded on the following seven indicators assessing ‘ eyes on the street ’ (for illustrations of se-lected built environment variables, see Figure 1 ):

1. Above grade: defi ned as buildings that “ sat ” at least 12 inches above the level of the sidewalk.

2. Stoop: a small raised platform at the entrance of a building, typically composed of several steps, which provides a place for sitting.

3. Porch: any covered exterior space protecting the entrance to a building.

4. Ground fl oor parking was defi ned when the ground fl oor of a multistory building was dedicated to parking (hypothesized to be detrimental).

5. Window area refl ects the proportion of building face comprising windows.

6. Low sill height windows (<3 feet from the sill of the dominant fi rst fl oor window to the main level of the fi rst fl oor). Low sill height allows the occupant to see more easily out to the street.

7. Setback refl ects the distance from the building to the street (smaller setback being preferred, in order to see out over the street). For analytic purposes, this reverse-scored variable is termed “ Small Setback. ”

Data regarding these seven features were then weighted according to the frontage of the lot (i.e., the proportion of the block face corresponding to each lot) and each feature aggregated for each of the 403 blocks in East Little Havana. The resulting variable was an estimate of the proportion of the total block frontage for which each built environment feature was present. For further detail regarding the UM-BECS coding system, see Lombard et al. (2000) and Spokane et al. (2007) .

Perceived social support was assessed by two indicators in the Spanish language from Resources for Enhancing Al-zheimer’s Caregiver Health ( Wisniewski et al., 2003 ), based on items from Krause and colleagues ( Krause, 1995 ; Krause & Markides, 1990 ). These indicators are highly related to each other and to elders ’ behavioral health ( Brown et al., in press ; Krause & Liang, 1993 ; Krause et al., 1989 ): (a) “ Satisfac-tion with support over the last month ” assesses the elder’s satisfaction with tangible, informational, and emotional sup-port received by her/him in the last month and consists of three, 4-choice Likert items (e.g., “ Overall, how satisfi ed have you been in the last month with the help you have re-ceived with transportation, housework, yard work, and shop-ping? ” ; for the present sample, a = .71) and (b) “ Satisfaction

Figure 1. Illustrations of selected built environment variables.

BROWN ET AL. 238

with support over the last year ” assesses the elder’s satisfac-tion with tangible, informational, and emotional support re-ceived by her/him in the last year and consists of three, 3-choice Likert items (e.g., “ Are you satisfi ed with the amount of emotional support that you have received from others, or do you wish that others gave you this type of help more often or less often? ” a = .73).

Psychological distress was measured by self-reported anxi-ety and depressive symptoms. Anxiety was measured using a 10-item, Spanish version of the Spielberger State Anxiety In-ventory, taken directly from Spielberger’s larger State-Trait Personality Inventory ( Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983 ). This measure has well-documented adequate reliability and validity (for this sample, a = .89) ( Carmin, Pollard, & Gillock, 1999 ; Mote, Natalicio, & Rivas, 1971 ; Novy, Nelson, Smith, & Rogers, 1995 ). Depressive symptoms were assessed using the Spanish version of the Center for Epidemiological Studies–Depression scale ( Carmin et al., 1999 ; Radloff, 1977 ), a 20-item, 4-choice Likert scale of symptoms that place an individual at risk for clinical depres-sion. The measure has demonstrated adequate test – retest reliability and is widely used with the elderly (for this sam-ple, a = .87) ( Carmin et al. ; Miller, Markides, & Black, 1997 ; Narrow, Rae, Moscicki, Locke, & Regier, 1990 ).

Analytic Strategy: Structural Equation Modeling The hypothesized relationships were tested using structural

equation modeling. Structural equation modeling was chosen over other analytic techniques, such as measured-variable path analysis, because it allows for investigation of complex models including relations among observed and latent vari-ables and corrects for measurement error, thereby allowing for a more accurate test of mediational effects ( Kaplan, 2000 ; Kline, 2005 ; Loehlin, 2004 ; McCoach, Black, & O’Connell, 2007 ). Structural equation modeling of the relationship of the built environment to perceived social support and psycho-logical distress was conducted using AMOS 6 ( Arbuckle & Wothke, 2005 ), while controlling for the relationships of age, gender, education, and income to psychological distress. Ad-ditionally, elders ’ global cognitive functioning (i.e., MMSE; Folstein et al., 1975 ) was free to correlate with the built envi-ronment, support, and distress variables. Any missing data were addressed using a full information maximum likelihood

algorithm ( Arbuckle, 1996 ); however, there was a low level of missing data (<1.4%). Further analyses then assessed whether perceived social support mediated the relationship between any of the built environment variables and psycho-logical distress. Mediation was tested using the asymmetrical distribution of products test (cf. MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002 ), which multiplies the un-standardized coeffi cients of the two paths that determine the mediating pathway and estimates a corresponding standard error. If the 95% confi dence interval for this product does not include zero, then mediation is assumed. This approach is used rather than the traditional approach of Baron and Kenny (1986) to test for mediation because it is a more powerful test and is more appropriate for structural equation models ( MacKinnon et al., 2002 ). Moreover, the asymmetrical distri-bution of products test has been shown to be superior to fur-ther methods of testing for mediation such as the distribution of the product of standardized path coeffi cients, both in terms of reducing the Type I error rate and in terms of being more accurate than methods based on an assumed normal distribu-tion ( MacKinnon et al., 2002 ; MacKinnon, Lockwood, & Williams, 2004 ).

Results

Preliminary Analyses Table 1 shows the zero-order correlations among the

seven built environment variables. The selection of specifi c built environment latent variables was based on an exami-nation of the covariance and correlation matrices for these variables because there were no prior data on which to base a priori predictions about how these data might be aggre-gated (i.e., this is the fi rst study to examine these ‘ eyes on the street ’ variables). Based on the corresponding covari-ances, a model was developed in which three of these vari-ables were combined into a latent variable, labeled “ Front Entrance, ” that refers to the relationship of the building’s entrance to the street and consisted of the proportion of a block in which buildings were above grade, possessed a stoop, and had a porch. Two other built environment vari-ables were combined into a second latent variable, labeled “ Windows, ” that refers to features that increase visibility from the interior of a building and consisted of the average

Table 1. Zero-Order Correlations Among Built Environment Variables

1 2 3 4 5 6 7

1 Grade — 2 Stoop 0.544** — 3 Porch 0.257** 0.346** — 4 Ground parking − 0.108 − 0.199** − 0.078 — 5 Window area 0.357** 0.353** 0.105 − 0.196** — 6 Sill 0.164** 0.251** 0.133* − 0.089 0.378** — 7 Small Setback − 0.169** − 0.154* − 0.121* − 0.024 − 0.028 0.222** —

Notes : 1, Building built above grade; 2, presence of a stoop; 3, presence of a porch; 4, in a multistory building, a ground fl oor that is devoted to parking; 5, window area; 6, windows with low sill height; 7, the distance the building is set back from the street (reverse scored so that higher values indicate smaller setbacks).

* p < .05; ** p < .01.

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 239

portion of the building face devoted to windows on the block and the proportion of the block corresponding to low sill height. The remaining two built environment variables, “ Ground-Floor Parking ” and “ Small Setback, ” had weaker and less consistent relationships with the other six built en-vironment variables and were therefore considered sepa-rately, as individual variables in the model (see Table 1 ).

In contrast, there existed an a priori plan to create spe-cifi c latent variables for perceived social support and psy-chological distress based on the multiple indicators of each, assuming the correlations among the corresponding indica-tor variables were suffi cient.

Perceived social support: The 1-month and 1-year mea-sures of support satisfaction were signifi cantly correlated ( r = .312, p < .01) and were therefore combined into a single latent variable of perceived social support.

Psychological distress: The measures of anxiety and de-pressive symptoms were signifi cantly correlated ( r = .773, p < .001) and were therefore incorporated into a single latent variable of psychological distress.

Table 2 presents the factor loadings and distributional characteristics of the indicators for each latent variable in the fi nal model, with factor loadings in the acceptable range ( Comrey & Lee, 1992 ). Using these latent variables, a pre-liminary model was developed in which the built environ-ment features predicted perceived social support and psychological distress, and perceived social support predicted psychological distress. In this model, the two built environ-ment latent variables, Front Entrance and Windows, and the two observed built environment variables, Ground-Floor Parking and Small Setback, were allowed to covary. The built environment variables were also allowed to predict perceived social support and psychological distress. Perceived social support in turn was allowed to predict psychological distress. These hypothesized relationships are depicted in Figure 2 .

Final Model Based on the preliminary model, we conducted a subse-

quent analysis which eliminated all nonsignifi cant pathways to derive a fi nal model (see Figure 3 for the structural portion of the model and Table 2 for the measurement portion of the model indicating the factor loadings of the individual indica-tors in the fi nal model). Eliminating the fi ve nonsignifi cant pathways (i.e., the one path from the Small Setback latent variable to perceived social support and the four paths from each of the four built environment variables to psychological distress) had no meaningful impact on the signifi cance or sta-bility of the other relationships tested. A chi-square difference test comparing the preliminary model and the fi nal model fur-ther revealed there was not a statistically signifi cant differ-ence in the fi t of the two models to the data ( c 2 (5) = 3.041, p > .69). Consequently, the simpler model with the fewest number of paths is preferred. Thus, only the results for the fi nal model as depicted in Figure 3 (all paths presented were statistically signifi cant, p < .05) are reported subsequently.

The fi t indices computed for the fi nal model suggested an acceptable fi t to the data, c 2 (87) = 119.106, p = .013, c 2 / df = 1.369, Comparative Fit Index (CFI) = 0.948, Root Mean Square Error of Approximation (RMSEA) = 0.037 ( Arbuckle, 2005 ; Kline, 2005 ). (Standard fi t indices for structural equation modeling were used to evaluate the fi t of the overall model to the data: Values of c 2 /df < 3, CFI > 0.90, and RMSEA < 0.08 suggest adequate model fi t [ Arbuckle, 2005 ; Kline, 2005 ].) In the fi nal model, the two latent built environment variables and the two observed built environment variables continued to be free to correlate with each other, as did the control variables. (With regard to signifi cant correlations between the control variables and the main analytic variables in the fi nal model, there was a signifi cant correlation between global cognitive functioning and perceived social support [ φ = 0.28], indicating that

Figure 2. Conceptual model of the relationship of built environment “ eyes on the street ” features to perceived social support and psychological distress. Dotted

lines indicate hypothesized indirect relationships (i.e., perceived social support is hypothesized to mediate the relationship of built environment variables to psycho-logical distress).

BROWN ET AL. 240

elders with higher levels of cognitive functioning reported being more satisfi ed with their social support. In addition, psychological distress was signifi cantly correlated with both income [ φ = − 0.14] and female gender [ φ = 0.20], sug-gesting that psychological distress was higher in lower-income elders compared with higher-income elders and was higher in women than in men.) Each of the two latent built environment variables predicted social support, albeit in dif-ferent directions. As predicted, the Front Entrance latent variable was predictive of higher levels of perceived social support, indicating that elders who lived on blocks with greater proportions of frontage that included porches, stoops, and buildings built above-grade reported higher scores on the social support variables. In contrast, the Windows latent variable was also related to perceived social support, but in

the opposite direction of what was predicted, indicating that elders who lived on blocks with greater proportions of win-dow area and/or greater low sill height reported lower levels of social support. With regard to the observed built environ-ment variables, Ground-fl oor Parking was negatively related to perceived social support, with blocks with higher propor-tions of ground-fl oor parking associated with lower levels of social support. The remaining observed built environment variable, Small Setback, was not related to social support. None of the built environment variables were directly re-lated to psychological distress. However, as predicted, the perceived social support latent variable was inversely related to the psychological distress latent variable, with those el-ders reporting higher levels of social support reporting lower levels of depressive symptoms and anxiety.

Table 2. Loadings and Distributional Statistics of Measured Indicators in the Final Structural Equation Model

VariableStandardized Lambda ( l ) Loadings Mean (Observed) ( SD ) (Observed) Range (Observed)

Built environment variables a Front entrance Above grade 0.69 0.44 0.25 0 – 1 Stoop 0.79 0.30 0.21 0 – 0.91 Porch 0.41 0.09 0.11 0 – 0.55 Ground fl oor parking — 0.02 0.06 0 – 0.50 Windows Window area 0.68 0.63 0.17 0.10 – 0.98 Low sill height 0.55 0.44 0.23 0 – 1 Small setback b — 0.25 0.18 0 – 0.99 Psychosocial variables Perceived social support Support satisfaction (month) 0.46 7.61 2.13 0 – 9 Support satisfaction (year) 0.73 5.66 0.89 3 – 9 Psychological distress Anxiety 0.81 20.33 8.10 10 – 40 Depressive symptoms 0.96 12.99 10.63 0 – 42

Notes : a Values indicate the proportions of block frontage that correspond to each built environment feature. b Small setback is reverse scored so that higher values correspond to shorter distances from the building to the street.

*Controls for gender, age, education, income, and functional status.

X2(95) = 130.779, p = .009X2/df = 1.369CFI = .948RMSEA = .037

SMALLSETBACK

-.51

.40

-.53

FRONTENTRANCE

PSYCHOLOGICALDISTRESS*

WINDOWS

PERCEIVEDSOCIAL SUPPORT

GROUND FLOORPARKING

-.22

Figure 3. Final structural equation model of the relationship of built environment “ eyes on the street ” features to perceived social support and psychological dis-

tress. Standardized beta estimates are reported for each path signifi cant at p < .05. Standard fi t indices are reported for the overall model, which suggest acceptable model fi t ( Kline, 2005 ).

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 241

Tests for Mediation Analyses then examined whether perceived social sup-

port mediated each of the three signifi cant relationships between the built environment (i.e., Front Entrance, Win-dows, and Ground-Floor Parking) and psychological dis-tress. First examined was whether perceived social support mediated the relationship between the Front Entrance latent variable and psychological distress. This was done by mul-tiplying the unstandardized path coeffi cient from the Front Entrance latent variable to perceived social support, by the unstandardized path coeffi cient from perceived social sup-port to psychological distress and estimating a correspond-ing standard error for this product term: The confi dence interval for this initial test of mediation did not include zero (product of unstandardized paths = – 8.09, 95% confi dence interval ranged from – 13.31 to – 2.88). Because zero is not included in the confi dence interval, perceived social support appears to mediate the relationship between Front Entrance and psychological distress. Second examined was whether perceived social support mediated the relationship from the Windows latent variable to psychological distress. Again, this was done by multiplying the unstandardized path coef-fi cient from the Windows latent variable to perceived social support, by the unstandardized coeffi cient from perceived social support to psychological distress, and estimating a corresponding standard error. The confi dence interval for this test once again did not include zero (product of unstan-dardized paths = 13.47, 95% confi dence interval ranged from 5.27 to 21.67). Therefore, perceived social support can be said to mediate the relationship between Windows and psychological distress. Third, we examined whether per-ceived social support mediated the relationship from ground-fl oor parking to psychological distress. The confi -dence interval for this test once again did not include zero (product of unstandardized paths = 11.71, 95% confi dence interval ranged from 3.88 to 19.53). Therefore, perceived social support can be said to mediate the relationship be-tween ground-fl oor parking and psychological distress.

Discussion As predicted, results suggested a two-step process, in

which features of the built environment impacted perceived social support, which subsequently impacted psychological distress. Additionally, we found that perceived social sup-port mediated the relationship between built environment variables and psychological distress, which is consistent with the view that social support may be one of the underly-ing mechanisms through which the built environment oper-ates on psychological distress ( Berke et al., 2007 ; Evans, 2003 ; Kruger et al., 2007 ; Thompson & Krause, 1998 ). Of particular interest is the differential direction of the relation-ships of the two latent built environment variables to per-ceived social support and psychological distress. As predicted, the Front Entrance latent variable, which included

porches, stoops, and buildings built above grade had a posi-tive relationship to perceived social support. In contrast, and contrary to prediction, the Windows latent variable, which included window area and low sill-height windows, had a negative relationship to perceived social support. It is pos-sible that features such as front porches, stoops, and homes built above grade, which provide opportunity for person-to-person contact and interaction, may encourage the elder to go outside and hence facilitate the direct social interactions leading to higher perceived levels of support. On the other hand, while larger window areas and lower sill heights may allow for broader observation of the surrounding area, they may also remove individuals from close person-to-person contact with elders and hence may result in reduced feelings of personal social support for elders themselves. Thus, for elders living in a neighborhood like East Little Havana, social support may be a “ ground-level ” phenomenon, facilitated by built environment features that promote direct, close-proximity, person-to-person interaction in the exterior of a building (see also, e.g., Kweon et al., 1998 ; Skjaeveland & Garling, 1997 ).

In addition, post hoc analyses found that buildings with porches and stoops tended to be older (pre-1945) than build-ings without porches and stoops (post-1945). One interpre-tation is that pre-1945 buildings are most likely to be associated with higher levels of perceived social support among neighbors and residents because they incorporate features such as porches and stoops that increase face-to-face interactions ( Katz, 1994 ; Leccese & McCormick, 2000 ) and related pedestrian activity ( Berrigan & Troiano, 2002 ). This interpretation is consistent with the fi nding that residents ’ levels of depressive symptoms were higher in dwellings of recent construction (i.e., post-1969 construc-tion) than in dwellings of less recent construction (i.e., pre-1969 construction) ( Weich et al., 2002 ), and the fi nding that elders ’ depressive symptoms were higher in less walkable neighborhoods than in more walkable neighborhoods, which are often older ( Berke et al., 2007 ; Handy, 1996 ). This interpretation is also consistent with recent sugges-tions that social support may mediate the impacts of the built environment on residents ’ mental health (e.g., Berke et al., 2007 ).

As predicted, higher levels of ground-fl oor parking on a block were associated with less perceived social support and greater psychological distress. This was not surprising, given that ground-fl oor parking (or a building built directly above a parking lot) is thought to be the equivalent of an obstructed front entrance, which effectively reduces visual and physical contacts between the resident in the building and the pedestrian on the street.

It was surprising, however, that the Windows latent vari-able had a negative relationship to perceived social support. It was predicted that living on a block where either a high proportion of the building face comprised windows or a greater proportion of windows had low sill heights would

BROWN ET AL. 242

lead to higher levels of perceived support because window area and lower sill heights are believed to allow residents inside the home to better observe and monitor events occurring outside. It is possible that when the building faces on a block are covered almost entirely by windows, residents on that block may feel less individual responsibility to assist in an emergency or nonemergency situation because they see that other observers are potentially available to provide help, and thus the responsibility for helping is diffused among multiple observers on that block ( Darley & Latané, 1968 ; Jones & Foshay, 1984 ). An alternative explanation may be that, due to residents ’ concerns about privacy, the presence of substantial window area on the block may re-sult in residents on that block feeling more “ exposed, ” and therefore residents may be more likely to block windows with curtains or blinds, thus inhibiting interaction. This may particularly be the case when extensive proportions of building faces are covered by glass (i.e., fl oor-to-ceiling windows), given that large expanses of glass do not allow the same type of privacy as a single, traditional window. This interpretation is consistent with recent suggestions that psychosocial functioning is worse when conditions do not promote a suffi cient balance between privacy and pub-lic space ( Skjaeveland & Garling, 1997 ; Weich et al., 2002 ). A feeling of vulnerability to outsiders or concern about se-curity may be particularly relevant to elders who reside in a low-SES urban neighborhood ( Thompson & Krause, 1998 ; Young, Russell, & Powers, 2004 ). A third possible explanation is that when most of the building faces on a block are covered with windows, elders may not go out-side, which may inhibit much needed social interaction: This is consistent with the longstanding view in the geron-tological literature that environmental conditions should strike a balance between needs for comfort and challenge to maintain optimal psychological functioning ( Lawton, 1989 , 1998 ).

With further regard to the negative relationship of the two Windows variables (i.e., window area and low sill height) to social support, we are not suggesting that it is optimal to have no windows on a block. That would be extrapolating beyond the data, given that all of the blocks in East Little Havana contained at least some windows (i.e., >10% of the building faces on each block consisted of windows). More-over, we acknowledge previous fi ndings suggesting that windows may have multiple health benefi ts, including the restorative effects of views to nature ( Ulrich, 1984 ) and of exposure to daylight from windows ( Küller & Lindsten, 1992 ) on individuals ’ health and well-being. Rather, we note that all elders had at least some windows, but that for this population of elders in this particular urban community of East Little Havana, blocks with the greatest proportions of window area and of windows with low sill height were associated with the least amounts of perceived social sup-port and greatest amounts of psychological distress among our elders.

Taken as a whole, the fi ndings suggest that the highest benefi ts for elders appear to occur on blocks with a greater number of positive front-entrance features (e.g., porches, stoops), which tend to be above grade and which offer a three-dimensional “ transitional space ” between the interior and the exterior of the building and between the height of the pedestrian and the height of the resident. This space is clearly the domain of the resident but beyond the privacy of the interior. A resident who chooses to occupy the transi-tional space just above the direct eye-to-eye level of the pe-destrian is announcing a willingness to participate in the life of the street without inviting the street into the house. There-fore, people looking at someone on a stoop or a porch are responding to a social invitation, in which the resident’s presence outside announces a willingness to interact or, at the very least, to be seen. This may account for the greater perceived social support and reduced psychological distress of residents who live on blocks with more porches and stoops, as compared with blocks with fewer porches and stoops (for similar fi ndings, see, e.g., Abu-Ghazzeh, 1999 ; Skjaeveland & Garling, 1997 ). In contrast, blocks where ei-ther a high proportion of the building face comprised win-dows or a greater proportion of windows had low sill heights may be associated with reductions in perceived social sup-port and greater psychological distress because these built environment features may promote unwanted visual/social contacts (i.e., people looking into someone’s home are un-bidden) and therefore the resident can only choose whether to be observed or hidden. These unwanted visual or social contacts may result in a loss of personal control, and much prior work suggests that uncontrollable social interaction may result in increased helplessness, greater stress, and feelings of psychological distress (see, e.g., Baum & Valins, 1979 ; Evans, 2003 ).

The present fi ndings build upon the authors ’ prior “ trans-disciplinary ” program of research — refl ecting the combined contributions of behavioral scientists, health scientists, and architects — and suggest that objective built environment fea-tures impact community social processes and residents ’ be-havioral health outcomes ( Spokane et al., 2007 ; Szapocznik et al., 2006 ). Moreover, these results are reminiscent of earlier fi ndings suggesting that built environment character-istics (e.g., green outdoor common spaces) infl uence social interactions and behavioral health (e.g., coping) in inner-city adults ( Kuo, 2001 ; Kweon et al., 1998 ; Sullivan, Kuo, & De Pooter, 2004 ).

Several strengths inherent in this study should be noted. This is one of the fi rst studies to explore both neighborhood physical and social conditions as possible protective factors for behavioral health problems in Hispanic elders residing in urban, low-SES neighborhoods (e.g., Espino, Lichten-stein, Palmer, & Hazuda, 2001 ). In addition, this is the fi rst study of which we are aware to examine block-level built environment variables as predictors of protection/risk in el-derly adults, based on a theory-driven model ( Duany, 2000 ;

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 243

Jacobs, 1992 ; Leccese & McCormick, 2000 ). In addition, this study captured fi ne-grained lot-level information on the built environment in the area immediately around the indi-vidual’s home, which may be especially applicable to elders rather than arbitrary Census-defi ned boundaries (see, e.g., Berke et al., 2007 ). Moreover, the study uses objective mea-sures of built environment and uses a population-based sample of community-dwelling Hispanic elders. The longi-tudinal nature of the data collection provides additional strength: Built environment data were collected in 2000 – 2002, whereas psychosocial data were collected in 2002 – 2004, allowing for greater confi dence in judgments of causality. However, this analytic strategy precluded us from examining the impacts of changes in the built environment, social support, and psychological distress over time. Future research should explore change trajectories among these variables over longer time periods. Finally, we identifi ed substantial variability in built environment which was pre-dictive of social support and psychological distress.

Nevertheless, the inherent inability to randomly assign elders to blocks allows for potential self-selection bias. For example, elders who are less sociable or more depressed or of lower SES may move to blocks that have fewer positive built environment features that encourage social interaction (e.g., porches, stoops). In addition, while built environment was assessed objectively, measures of social support and psychological distress were self-reported, and unmeasured personality characteristics or attributional styles may have impacted those measures. Similarly, additional unmeasured mediating variables such as collective effi cacy, block-level SES, block-level density, residential stability, safety, or age distribution of residents on each block may have also ac-counted for variance in elders ’ perceived social support and psychological distress ( Cutrona, Wallace, & Wesner, 2006 ; Espino et al., 2001 ; Galea, Ahern, Rudenstine, Wallace, & Vlahov, 2005 ; Kubzansky et al., 2005 ; Sampson, Rauden-bush, & Earls, 1997 ; Thompson & Krause, 1998 ). Addition-ally, the present research examined effects of the built environment at the block level, and future research should consider the possible mental health impacts of the built en-vironment at other geographic levels of analysis (e.g., the home level, the neighborhood level). Similarly, we did not assess whether participants owned their own home, although there was a low rate of home ownership (8.6%) for East Little Havana as a whole ( U.S. Census Bureau, 2000 ). Moreover, although we considered time in home as an ad-ditional covariate in our analyses, this was unrelated to the psychosocial variables or to any of the built environment variables — which were examined at the block level. Future research should therefore consider the potential impacts of home ownership, time in home, and building characteristics on elders ’ social and mental health outcomes, over and be-yond the effects of surrounding neighborhood environmen-tal characteristics (see, e.g., Brown, Perkins, & Brown, 2003 ; Krause, 1998 ; Lawton, Nahemow, & Teaff, 1975 ;

Young et al., 2004 ). A further limitation is that over 40% of older adults randomly selected for participation were not enrolled, primarily due to deaths, refusals, and moves away from the study area. Furthermore, the work is based on el-ders, ages 70 and above, in a single community with its own specifi c characteristics, such as low SES, predominantly Hispanic, with limited home ownership ( U.S. Census Bu-reau, 2000 ), and with purely residential blocks within a 5-min walk of commercial blocks, which may not be refl ec-tive of other communities in which elders live. Similarly, although we titled the manuscript “ Hispanic, ” we acknowl-edge that 87% of the sample was Cuban born. This may limit the generalizability of the fi ndings, requiring replica-tion in both similar and fundamentally different communi-ties, including other Hispanic subgroups and non-Hispanics, and with larger sample sizes. Finally, although our fi ndings are consistent with prior suggestions that built environment features may promote social contacts with neighbors ( Jacobs, 1992 ) and perceived availability of support ( Thompson & Krause, 1998 ), our social support items did not specify the referent for the social support transaction (e.g., whether the source of any received support was a friend or a family member or neighbor of the elder). Future research on the relationship of the built environment to social support should use more sensitive measures of the source, type, and reaction to social support.

In sum, the results suggest that the neighborhood built environment in which Hispanic elders live may have a mod-est, but important positive impact on their perceptions of the quality and quantity of social support that they receive, which can then lead to reduced psychological distress. From a public health perspective, even these “ small ” effects of the built environment can be of considerable signifi cance be-cause of the pervasiveness of the built environment. The fi nding of this and other studies suggesting the neighbor-hood built environment may be an important determinant of health ( Berke et al., 2007 ; Clarke & George, 2005 ; Diez Roux, 2003 ; Evans, 2003 ; Frank et al., 2004 ; Institute of Medicine and Transportation Research Board, 2005 ; Kruger et al., 2007 ; Rohrer et al., 2004 ; Szapocznik et al., 2006 ) suggests the need for more research to identify the specifi c environmental characteristics that may best promote elders ’ mental functioning and independence.

Funding

This work was supported by a National Institute of Mental Health/Na-tional Institute of Environmental Health Sciences Grant No. MH 063709 (J.S., Principal Investigator [PI]; A.R.S., Co-Principal Investigator [Co-PI]), by a National Institute on Aging Grant No. AG 027527 (J.S., PI; S.C.B., Co-PI), and by a Robert Wood Johnson Foundation Grant No. 037377 (J.S., PI; E.P.-Z, Co-PI).

Acknowledgments

We acknowledge Monica Zarate, Rosa Verdeja, Tatiana Clavijo, Aleyda Marcos, and Patricia Thomas for their assistance in the conduct of the study. We thank Dr. Tatiana Perrino for her helpful comments on an earlier version of the manuscript. S.C.B. was responsible for the conduct of the

BROWN ET AL. 244

study, wrote and revised the manuscript, supervised data analysis and interpretation, and performed data analysis and interpretation. C.A.M. performed data analysis and interpretation, provided statistical and methodological expertise, and assisted with writing and revising the manu-script. J.L.L. and F.M. supervised the collection of the built environment data and contributed to writing and revising the manuscript and interpret-ing the analytic results. E.P.-Z contributed to study conceptualization and interpretation of the analytic results. A.R.S. and F.L.N. contributed to study conceptualization, supervising the conduct of the study, writing and revis-ing the manuscript, and interpreting the analytic results. H.P. contributed to writing and revising the manuscript and interpreting the analytic results. J.S. conceived of the study and supervised all aspects of its implementa-tion, including writing and revising the manuscript and interpreting the analytic results.

Correspondence

Address correspondence to Scott C. Brown, PhD, Department of Epide-miology and Public Health, Center for Family Studies, University of Mi-ami Miller School of Medicine, 1420 NW 10th Ave., Miami, FL 33136. E-mail: sbrown@med.miami.edu

References Abu-Ghazzeh , T. M. ( 1999 ). Housing layout, social interaction, and the

place of contact in Abu-Nuseir, Jordan . Journal of Environmental Psychology , 19 , 41 – 73 .

Adler , N. , & Matthews , K. ( 1994 ). Health psychology: Why do some peo-ple get sick and some stay well? Annual Review of Psychology , 45 , 229 – 259 .

Arbuckle , J. L. ( 1996 ). Full information estimation in the presence of in-complete data . In G. A. Marcoulides & R. E. Schumacker (Eds.), Advanced structural equation modeling: Issues and techniques (pp. 243 – 277 ). Mahwah, NJ : Erlbaum .

Arbuckle , J. L. ( 2005 ). AMOS 6.0 user’s guide . Spring House, PA : AMOS Development Corp .

Arbuckle , J. L. , & Wothke , W. ( 2005 ). AMOS release 6.0 [computer pro-gram] . Chicago : Smallwaters .

Baron , R. M. , & Kenny , D. A. ( 1986 ). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations . Journal of Personality and Social Psy-chology , 51 , 1173 – 1182 .

Baum , A. , & Valins , S. ( 1979 ). Architectural mediation of residential den-sity and control: Crowding and the regulation of social contact . In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 131 – 175 ). New York : Academic .

Berke , E. M. , Gottlieb , L. M. , Vernez Moudon , A. , & Larson , E. B. ( 2007 ). Protective association between neighborhood walkability and de-pression in older men . Journal of the American Geriatrics Society , 55 , 526 – 533 .

Berrigan , D. , & Troiano , R. P. ( 2002 ). The association between urban form and physical activity in U.S. adults . American Journal of Preventive Medicine , 23 ( 2S ), 74 – 79 .

Brown , B. , Perkins , D. D. , & Brown , G. ( 2003 ). Place attachment in a re-vitalizing neighborhood: Individual and block levels of analysis . Journal of Environmental Psychology , 23 , 259 – 271 .

Brown , S. C. , Mason , C. A. , Spokane , A. R. , Cruza-Guet , M. C. , Lopez , B. , & Szapocznik , J. (in press) . The relationship of neighborhood cli-mate to perceived social support and mental health in older Hispanic immigrants in Miami, FL . Journal of Aging and Health .

Cantor , M. H. ( 1979 ). Neighbors and friends: An overlooked resource in the informal support system . Research on Aging , 1 , 434 – 463 .

Carmin , C. N. , Pollard , C. A. , & Gillock , K. L. ( 1999 ). Assessment of anxiety disorders in the elderly . In P. A. Lichtenberg (Ed.), Handbook of assessment in clinical gerontology (pp. 59 – 90 ). New York : Wiley .

Chou , K. L. ( 1999 ). Social support and subjective well-being among Hong Kong Chinese young adults . Journal of Genetic Psychology , 160 , 319 – 331 .

Clarke , P. , & George , L. K. ( 2005 ). The role of the built environment in the disablement process . American Journal of Public Health , 95 ( 11 ), 1933 – 1939 .

Cohen , S. ( 2004 ). Social relationships and health . American Psychologist , 59 , 676 – 684 .

Cohen , S. , & Wills , T. A. ( 1985 ). Stress, social support and the buffering hypothesis . Psychological Bulletin , 98 , 310 – 357 .

Comrey , A. L. , & Lee , H. B. ( 1992 ). A First course in factor analysis (2nd ed.) . Hillsdale, NJ : Lawrence Erlbaum Associates .

Cornman , J. C. , Goldman , N. , Glei , D. A. , Weinstein , M. , & Chang , M. C. ( 2003 ). Social ties and perceived support: Two dimensions of social relationships and health among the elderly in Taiwan . Journal of Ag-ing and Health , 15 ( 4 ), 616 – 644 .

Cutrona , C. E. , Wallace , G. , & Wesner , K. A. ( 2006 ). Neighborhood char-acteristics and depression . Current Directions in Psychological Sci-ence , 15 , 188 – 192 .

Darley , J. M. , & Latané , B. ( 1968 ). Bystander intervention in emergencies: Diffusion of responsibility . Journal of Personality and Social Psy-chology , 8 , 377 – 383 .

Diez Roux , A. V. ( 2003 ). Residential environments and cardiovascular risk . Journal of Urban Health , 80 , 569 – 589 .

Duany , A. ( 2000 ). A new theory of urbanism . Scientifi c American , 28 ( 6 ), 90 – 91 .

Espino , D. V. , Lichtenstein , M. J. , Palmer , R. F. , & Hazuda , H. P. ( 2001 ). Ethnic differences in Mini-Mental State Examination (MMSE) scores: Where you live makes a difference . Journal of the American Geriatrics Society , 49 , 538 – 548 .

Evans , G. W. ( 2003 ). The built environment and mental health . Journals of Urban Health , 80 , 536 – 555 .

Falcón , L. M. , & Tucker , K. L. ( 2000 ). Prevalence and correlates of depres-sive symptoms among Hispanic elders in Massachusetts . Journals of Gerontology: Social Sciences , 55 ( 2 ), S108 – S116 .

Fleming , R. , Baum , A. , & Singer , J. E. ( 1985 ). Social support and the phys-ical environment . In S. Cohen & L. Syme (Eds.), Social support and health (pp. 327 – 346 ). New York : Academic .

Folstein , M. F. , Folstein , S. E. , & McHugh , P. R. ( 1975 ). Mini-mental state: A practical method for grading the cognitive state of patients for the clinician . Journal of Psychiatric Research , 12 , 89 – 198 .

Frank , L. D. , Andresen , M. A. , & Schmid , T. L. ( 2004 ). Obesity relation-ships with community design, physical activity, and time spent in cars . American Journal of Preventive Medicine , 27 , 87 – 96 .

Frumkin , H. , Frank , L. , & Jackson , R. ( 2004 ). Urban sprawl and public health: Designing, planning, and building for healthy communities . Washington, DC : Island Press .

Galea , S. , Ahern , J. , Rudenstine , S. , Wallace , Z. , & Vlahov , D. ( 2005 ). Urban built environment and depression: A multi-level analysis . Journal of Epidemiology and Community Health , 59 , 822 – 827 .

Handy , S. ( 1996 ). Understanding the link between urban form and non-work travel behavior . Journal of Planning Education and Research , 15 , 183 – 198 .

Horgas , A. L. , Wilms , H. U. , & Baltes , M. M. ( 1998 ). Daily life in very old age: Everyday activities as expression of successful living . Geron-tologist , 38 , 556 – 568 .

House , J. S. , Landis , K. R. , & Umberson , D. ( 1988 ). Social relationships and health . Science , 241 , 540 – 545 .

Institute of Medicine and Transportation Research Board . ( 2005 ). Does the built environment infl uence physical activity? Examining the evidence [Institute of Medicine and Transportation Research Board Report No. 282] . Washington, DC : National Academies Press .

Jacobs , J. ( 1992 ). The Death and life of great American cities . New York : Vintage Books . [Original work published in 1961] .

Jang , Y. , Haley , W. E. , Small , B. J. , & Mortimer , J. A. ( 2002 ). The role of mastery and social resources in the associations between disability and depression in later life . The Gerontologist , 42 , 807 – 813 .

BUILT ENVIRONMENT AND PSYCHOLOGICAL DISTRESS 245

Jones , L. M. , & Foshay , N. N. ( 1984 ). Diffusion of responsibility in a non-emergency situation: Response to a greeting from a stranger . The Journal of Social Psychology , 123 , 155 – 158 .

Kaplan , D. ( 2000 ). Structural equation modeling: Foundations and exten-sions . Thousand Oaks, CA : Sage .

Katz , P. ( 1994 ). The New Urbanism: Toward an architecture of commu-nity . New York : McGraw-Hill .

Kawachi , I. ( 1999 ). Social capital and community effects on population and individual health . Annals of the New York Academy of Sciences , 896 , 120 – 130 .

Kim , J. , & Kaplan , R. ( 2004 ). Physical and psychological factors in sense of community: New Urbanist Kentlands and nearby Orchard Village . Environment and Behavior , 36 , 313 – 340 .

Kline , R. B. ( 2005 ). Principles and practice of structural equation model-ing (2nd ed.). New York : Guilford .

Krause , N. ( 1995 ). Negative interaction and satisfaction with social sup-port among older adults . Journal of Gerontology: Psychological Sci-ences , 50 , P59 – P73 .

Krause , N. ( 1996 ). Neighborhood deterioration and self-rated health in later life . Psychology and Aging , 13 , 58 – 68 .

Krause , N. ( 1998 ). Neighborhood deterioration, religious coping, and changes in health during late life . The Gerontologist , 3 , 653 – 664 .

Krause , N. ( 1993 ). Neighborhood deterioration and social isolation in later life . International Journal of Aging and Human Development , 36 , 9 – 38 .

Krause , N. , Liang , J. ( 1993 ). Stress, social support, and psychological dis-tress among the Chinese elderly . Journal of Gerontology: Psycho-logical Sciences , 48 , P282 – P291 .

Krause , N. , Liang , J. , & Yatomi , N. ( 1989 ). Satisfaction with social support and depressive symptoms: A panel analysis . Psychology and Aging , 4 , 88 – 97 .

Krause , N. , & Markides , K. ( 1990 ). Measuring social support among older adults . International Journal of Aging and Human Development , 30 , 37 – 53 .

Kruger , D. J. , Reischl , T. M. , & Gee , G. C. ( 2007 ). Neighborhood social conditions mediate the association between physical deterioration and mental health . American Journal of Community Psychology , 40 , 261 – 271 .

Kubzansky , L. D. , Subramanian , S. V. , Kawachi , I. , Fay , M. E. , Soobader , M. J. , & Berkman , L. ( 2005 ). Neighborhood contextual infl uences on depressive symptoms in the elderly . American Journal of Epidemiol-ogy , 162 , 253 – 260 .

Küller , R. , & Lindsten , C. ( 1992 ). Health and behavior of children in class-rooms with and without windows . Journal of Environmental Psy-chology , 12 , 305 – 317 .

Kuo , F. E. ( 2001 ). Coping with poverty: Impacts of environment and atten-tion in the inner city . Environment and Behavior , 33 , 5 – 34 .

Kweon , B.-S. , Sullivan , W. C. , & Wiley , A. R. ( 1998 ). Green common spaces and the social integration of inner-city older adults . Environ-ment and Behavior , 30 ( 6 ), 832 – 858 .

Lawton , M. P. ( 1989 ). Environmental proactivity and affect in older peo-ple . In S. Spacapan , S. Oskamp (Ed.), The social psychology of aging (pp. 135 – 163 ). Newbury Park, CA : Sage .

Lawton , M. P. ( 1998 ). Environment and aging: Theory revisited . In R. J. Scheidt , & P. G. Windley (Eds.), Environment and aging theory: A focus on housing (pp. 1 – 31 ). Westport, CT : Greenwood Press .

Lawton , M. P. , Nahemow , L. , & Teaff , J. ( 1975 ). Housing characteristics and the well-being of elderly tenants in federally assisted housing . Journal of Gerontology , 30 , 601 – 607 .

Leccese , M. , & McCormick, K. (Eds.) ( 2000 ). Charter of the New Urban-ism . New York : McGraw-Hill .

Leyden , K. M. ( 2003 ). Social capital and the built environment: The im-portance of walkable neighborhoods . American Journal of Public Health , 93 , 1546 – 1551 .

Liang , J. , Krause , N. M. , & Bennett , J. M. ( 2001 ). Social exchange and well-being: Is giving better than receiving? Psychology and Aging , 16 , 511 – 523 .

Loehlin , J. C. ( 2004 ). Latent variable models: An introduction to factor, path, and structural equation analysis (4th ed.). Mahwah, NJ : Lawrence Erlbaum Associates .

Lombard , J. L. , Martinez , F. , Mason , C. A. , Szapocznik , J. , Gorman-Smith , D. , & Plater-Zyberk , E. ( 2000 ). University of Miami Built Environment Coding System Manual (UMBECS) . Retrieved September 30, 2008, from http://cfs.med.miami.edu/docs/publications/Built_Environment_Coding_Manual.pdf .

Lubben , J. E. ( 1988 ). Assessing social networks among elderly popula-tions . Family and Community Health , 11 , 42 – 52 .

MacKinnon , D. P. , Lockwood , C. M. , Hoffman , J. M. , West , S. G. , & Sheets , V. ( 2002 ). A comparison of methods to test mediation and other inter-vening variable effects . Psychological Methods , 7 , 83 – 104 .

MacKinnon , D. P. , Lockwood , C. M. , & Williams , J. ( 2004 ). Confi dence limits for the indirect effect: Distribution of the product and resam-pling methods . Multivariate Behavioral Research , 39 , 99 – 128 .

McCoach , D. B. , Black , A. C. , & O’Connell , A. A. ( 2007 ). Errors of infer-ence in structural equation modeling . Psychology in the Schools , 44 , 461 – 470 .

McCoy , J. M. ( 2002 ). Work environments . In R. B. Bechtel , & A. Churchman (Eds.), Handbook of environmental psychology ( pp. 443 – 460 , 2nd ed. ). New York : Wiley .

Miller , T. Q. , Markides , K. S. , & Black , B. A. ( 1997 ). The factor structure of the CES-D in two surveys of elderly Mexican Americans . Journal of Gerontology: Social Sciences , 52 , S259 – S269 .

Moos , R. H. ( 1976 ). The human context . New York : Wiley . Mote , T. A. , Natalicio , L. F. , & Rivas , F. ( 1971 ). Comparability of the

Spanish and English versions of the Spielberger State-Trait Anxiety Inventory . Journal of Cross-Cultural Psychology , 2 , 205 – 206 .

Narrow , W. E. , Rae , D. S. , Moscicki , E. K. , Locke , B. Z. , & Regier , D. A. ( 1990 ). Depression among Cuban Americans: The Hispanic Health and Nutrition Examination Survey . Social Psychiatry and Psychiat-ric Epidemiology , 25 , 260 – 268 .

Newsom , J. T. , & Schulz , R. ( 1996 ). Social support as a mediator in the relation between functional status and quality of life in older adults . Psychology and Aging , 11 , 34 – 44 .

Northridge , M. E. , Sclar , E. D. , & Biswas , P. ( 2003 ). Sorting out the con-nections between the built environment and health: A conceptual framework for navigating the pathways and planning healthy cities . Journal of Urban Health , 80 , 556 – 568 .

Novy , D. M. , Nelson , D. V. , Smith , K. G. , & Rogers , P. A. ( 1995 ). Psycho-metric comparability of the English and Spanish language versions of the State-Trait Anxiety Inventory . Hispanic Journal of Behavioral Sciences , 17 , 209 – 224 .

Peirce , R. S. , Frone , M. R. , Russell , M. , Cooper , M. L. , & Mudar , P. ( 2000 ). A longitudinal model of social contact, social support, depression, and alcohol use . Health Psychology , 19 , 28 – 38 .

Perrino , T. , Brown , S. C. , Mason , C. A. , & Szapocznik , J. (in press) . Depressive symptoms and urban Hispanic older adults in Miami: Prevalence and sociodemographic correlates . The Clinical Gerontologist .

Plehn , K. , Marcopulos , B. A. , & McLain , C. A. ( 2004 ). The relationship between neuropsychological test performance, social functioning, and instrumental activities of daily living in a sample of rural older adults . The Clinical Neuropsychologist , 18 , 101 – 113 .

Radloff , L. S. ( 1977 ). The CES-D Scale: A self-report depression scale for research in the general population . Applied Psychological Measure-ment , 1 , 385 – 401 .

Rohrer , J. , Pierce , J. R. , Jr. , & Denison , A. ( 2004 ). Walkability and self-rated health in primary care patients . BMC Family Practice , 5 , 29 .

Sampson , R. J. , Raudenbush , S. W. , & Earls , F. ( 1997 ). Neighborhoods and violent crime: A multilevel study of collective effi cacy . Science , 277 , 918 – 924 .

Seeman , T. E. , & Berkman , L. F. ( 1988 ). Structural characteristics of social networks and their relationship with social support in the elderly: Who provides support? . Social Science & Medicine , 26 , 737 – 749 .

BROWN ET AL. 246

Shaw , B. ( 2005 ). Anticipated support from neighbors and physical func-tioning during later life . Research on Aging , 27 , 503 – 525 .

Skjaeveland , O. , & Garling , T. ( 1997 ). Effects of interactional space on neighboring . Journal of Environmental Psychology , 17 , 181 – 198 .

Spielberger , C. D. , Gorsuch , R. L. , Lushene , P. R. , Vagg , P. R. , & Jacobs , G. A. ( 1983 ). State-Trait Anxiety Inventory for Adults (STAIS-AD) manual . Palo Alto, CA : Consulting Psychologists Press .

Spokane , A. R. , Lombard , J. L. , Martinez , F. , Mason , C. A. , Gorman-Smith , D. , Plater-Zyberk , E. , Brown , S. C. , Perrino , T. , & Szapoc-znik , J. ( 2007 ). Identifying streetscape features signifi cant to well-being . Architectural Science Review , 50 , 234 – 245 .

Stewart , A. L. , & Nápoles-Springer , A. ( 2000 ). Health-related quality of life assessments in diverse population groups in the United States . Medical Care , 38 , 125 – 129 .

Sullivan , W. C. , Kuo , F. E. , & De Pooter , S. F. ( 2004 ). The fruit of urban nature: Vital neighborhood spaces . Environment and Behavior , 36 , 678 – 700 .

Szapocznik , J. , Lombard , J. , Martinez , F. , Mason , C. A. , Gorman-Smith , D. , Plater-Zyberk , E. , Brown , S. C. , & Spokane , A. ( 2006 ). The im-pact of the built environment on children’s school grades: The role of diversity of use in a Hispanic neighborhood . American Journal of Community Psychology , 38 , 299 – 310 .

Thompson , E. E. , & Krause , N. ( 1998 ). Living alone and neighborhood characteristics as predictors of social support in late life . Journal of Gerontology: Social Sciences , 53 , S354 – S364 .

Ulrich , R. S. ( 1984 ). View through a window may infl uence recovery from surgery . Science , 224 , 420 – 421 .

U.S. Census Bureau . ( 2000 ). United States Census 2000 . Retrieved De-cember 30, 2008, from: http://www.census.gov/main/www/cen2000.html .

U.S. Census Bureau . ( 2002 ). Percent of people below poverty level. 2002 . Retrieved September 30, 2008, from http://www.census.gov/acs/www/Products/Ranking/2002/pdf/R01T160.pdf .

Weich , S. , Blanchard , M. , Prince , M. , Burton , E. , Erens , B. , & Sproston , K. ( 2002 ). Mental health and the built environment: Cross-sectional sur-vey of individual and contextual risk factors for depression . British Journal of Psychiatry , 180 , 428 – 433 .

Wellman , B. , Craven , P. , Whitaker , M. , du Toit , S. , Stevens , H. , & Bakker , H. ( 1973 ). Community ties and support systems: From intimacy to support . In L. Bourne , R. MacKinnon , J. Simmons (Eds.), The form of cities in central Canada: Selected papers (pp. 152 – 167 ). Toronto, Canada : University of Toronto Press .

Wethington , E. , & Kavey , A. ( 2000 ). Neighboring as a form of social inte-gration and social support . In K. Pillemer , P. Moen , & N. Glasgow (Eds.), Social integration in the second half of life (pp. 190 – 210 ). Baltimore : Johns Hopkins University Press .

Wisniewski , S. R. , Belle , S. H. , Coon , D. W. , Marcus , S. M. , Ory , M. G. , Burgio , L. D. , Burns , R. , & Schulz , R. ( 2003 ). The Resources for Enhancing Alzheimer’s Caregiver Health (REACH): Project design and baseline characteristics . Psychology and Aging , 18 , 375 – 384 .

Young , A. F. , Russell , A. , & Powers , J. R. ( 2004 ). The sense of belonging to a neighborhood: Can it be measured and is it related to health and well being in older women? . Social Science & Medicine , 59 , 2627 – 2637 .

Zimmer , Z. , Ofstedal , M. B. , & Chang , M.-C. ( 2001 ). Impact of cognitive status and decline on service and support utilization among older adults in Taiwan . Research on Aging , 23 , 267 – 303 .

Received February 5, 2008 Accepted October 8 , 2008 Decision Editor: Kenneth F. Ferraro, PhD